Applications that leverage natural language processing (NLP) have begun to achieve close to human-level accuracy in tasks such as language translation, text summarization, and text-to-speech, due to the adoption of deep learning models. This widespread adoption has been driven by two key developments in the area of deep learning. One of them is the rapid progress in discovering novel deep neural network architectures, realized by the availability of huge volumes of data. Such architectures can achieve superior performance compared to traditional approaches. The other development is the increasing availability of open source tools or libraries, such as TensorFlow, which make easy implementations of these modern architectures possible in practical or productive applications. The purpose of this chapter is to equip the reader with a necessary basic knowledge of deep learning and TensorFlow, so that later chapters can be approached with confidence.

The following topics will be covered in this chapter:

• Various concepts and terminologies in deep learning
• An overview of common deep learning architectures, such as CNNs and RNNs
• Installation, setup, and getting started with TensorFlow